INTRODUCTION: Primary gastric diffuse large B-cell lymphoma (PG-DLBCL) is a rare and heterogeneous entity within extranodal non-Hodgkin lymphomas. Due to its low prevalence, data on clinical characteristics, treatment patterns, and survival outcomes in real-world settings, particularly in Latin America, remain limited. This study aimed to describe the clinical features and treatment outcomes of patients with PG-DLBCL and to evaluate prognostic factors associated with overall survival (OS).

METHODS: We conducted a multicenter, retrospective cohort study using data from the GELL registry (2015–2024), which includes adult patients diagnosed with primary gastric diffuse large B-cell lymphoma (PG-DLBCL) and treated at academic hematology centers across Latin America. Consecutive patients who received first-line immunochemotherapy with R-CHOP, R-miniCHOP, or similar rituximab-based anthracycline regimens were included. Patients lost to follow-up were excluded. Baseline clinical and laboratory variables were collected at diagnosis. The primary endpoint was overall survival (OS), defined as the time from diagnosis to death from any cause or last known follow-up. The secondary endpoint was event-free survival (EFS), defined as the time from diagnosis to progression, relapse, unplanned treatment change, or death. The median follow-up time was estimated using the reverse Kaplan–Meier method. Missing data (<10%) were addressed via multiple imputation using Random Forest algorithms (mice R package). OS and EFS were estimated using the Kaplan–Meier method, and compared using the log-rank test. Multivariable Cox proportional hazards models were used to identify independent prognostic factors for OS and EFS. Predefined subgroup analyses were performed according to Helicobacter pylori (HP) status.

RESULTS: A total of 157 patients were included. The median follow-up time was 35.5 months (Interquartile Range: 13.1–68.6) and 50 (32%) died during follow-up. The median age of 64 years and a slight male predominance (53%). Most patients had ECOG ≤1 (92%), no B symptoms (52%), and localized disease (55%). R-CHOP was the most common first-line regimen (80%), with ≥6 cycles completed in 61% of cases. Based on NCCN-IPI, 46% had high-intermediate or high-risk scores (≥3). Relapse occurred in 12%, and overall mortality was 32%, mainly due to disease progression and sepsis. Most patients received R-CHOP or equivalent first-line therapy. Radiotherapy was used in 9.5% of deceased and 12.0% of surviving patients. OS at 36 months was 70% (95% CI: 62–78. EFS at 36 months was 80% (95% CI: 73–88. In multivariable analysis, age (HR: 1.05 per year; 95% CI: 1.02–1.07; p < 0.001), ECOG ≥2 (HR: 7.32; 95% CI: 3.09–17.3; p < 0.001), and advanced stage (HR: 2.57; 95% CI: 1.27–5.17; p = 0.008) were independently associated with increased mortality. For EFS, older age (HR = 1.03; 95% CI: 1.00–1.06; p = 0.033), ECOG ≥2 (HR = 12.3; 95% CI: 4.23–35.9; p < 0.001), and advanced stage (HR = 4.95; 95% CI: 1.62–15.2; p = 0.005) were independently associated with worse outcomes. In a subanalysis of 60 patients with known Helicobacter pylori status, HP-positive patients had significantly higher overall survival at 12 and 24 months (91.7% at both timepoints) compared to HP-negative patients (60.7% at 12 months and 56.7% at 36 months) and higher complete response rates (61% vs. 33%). Complete response was more common in HP-positive patients (46% vs. 26%). Disease progression (22% vs. 8%), relapse (2% vs. 0%), and treatment-related mortality (9% vs. 0%) were more frequent among HP-negative patients compared to those who were HP-positive.

Conclusions In this Latin American cohort, PG-DLBCL patients showed favorable survival outcomes with standard immunochemotherapy. Advanced age, poor performance status, and advanced stage were independently associated with worse OS and EFS. Helicobacter pylori positivity was linked to better survival, higher complete response rates, and fewer adverse outcomes, suggesting its potential role as a favorable prognostic factor in PG-DLBCL.

This content is only available as a PDF.
2025
Sign in via your Institution